Following the publication of the EU-wide stress test results by the European Banking Authority today, the Association for Financial Markets in Europe (AFME) issued a statement on behalf of its members that participated in the exercise.

Caroline Liesegang, Head of Prudential Regulation at AFME, said: “AFME welcomes the results of this year’s EBA stress tests. Despite an extreme ‘adverse scenario’ including high and persistent inflation and a severe decline in EU GDP, plummeting by 6% over the 3-year period, this year’s stress test shows that the steps that both banks and supervisors have taken over the years to strengthen the resilience of the EU banking sector are now paying off.

“The results reflect a better starting point for banks, with higher levels of capital, improved asset quality and profitability driving the change compared to the previous stress test. We also note that EU subsidiaries of international banks, included in the EU-wide stress test for the first time this year, have finished the exercise showing a robust solvency position.

“Notwithstanding the overall positive outcome, we urge the EBA to take a fresh look at the stress test methodology and remove or at least recalibrate some of the existing constraints that often override banks’ bottom-up projections. The EBA stress test follows a constrained bottom-up approach, involving banks in identifying risks using their own models to encourage better risk management practices. A successful stress test should find a balance between supervisory standardisation and accommodating individual bank characteristics.

“Finally, the new banking package (CRR3/CRD6) - the entry into force of which is expected by January 2025 - will warrant a comprehensive review of the EU stress test framework. The structural changes to the calibration of the Pillar 1 framework combined with other overlaps across the Pillar 1 and Pillar 2 capital risk coverage warrant a critical assessment of the methodology.

“We look forward to working with the EBA in further evolving the methodology for the 2025 stress test.”

In particular:

• Compared to the previous exercise, with an average depletion of 459 bps under the adverse scenario and the average the CET1 ratio in 2025 at 10.4% the results are an improvement to the 2021 exercise, under which the average capital depletion was 497bps, and the end-state CET1 ratio under the adverse scenario was 10.2%.

• Net interest income is the largest contributor to the increase in earnings (938 bps) and capital over the stress period, even if capped at the level of the starting point.

• Credit risk losses are the main negative contributor, with the increase in losses under the severely adverse economic scenario contributing  405 bps depletion to the CET1 ratios. It is also worth mentioning that market risk losses under the adverse scenario contribute 112 bps to the depletion in CET 1 ratios.

• As a new feature of the 2023 EU-wide stress test, net fees and commission income (NFCI) projections were prescribed to banks based on a centralised top-down model. Under the adverse scenario, the aggregate NFCI and dividend income decline by 22% on average over the three years of the horizon.

• In terms of wholesale activities, net trading income (NTI) drops significantly during the first year of the adverse scenario, marking a loss of 55bn EUR. The main drivers of the NTI drop are losses from positions in economic hedges, held with a trading intent (HFT) and liquidity reserves. HFT losses that were floored for banks’ projections amount to 21.5bn EUR (-25 bps), while client revenues in 2023 dropped by 44% (from 36bn to 20bn EUR), providing still a positive cumulative contribution to the NTI in the three years of the adverse scenario.

• Elsewhere, we note that the one-size-fits-all nature of the methodological constraints may severely impact banks’ final results based on business model rather than risk profile. The constraints in the methodology have a significant impact on specific areas of banks depending on bank specific business models. In particular, this year’s stress tests resulted in severe shocks for markets and securitisations businesses. Furthermore, with limited modelling allowed in net interest income and a uniform methodology applied to different business models, national and business model specificities with regard to net interest income may are not reflected in the outcome of the stress tests.

• Given the constraints, in many cases the bank internal view of risks and the impact of the stress scenario is not in line with the final stress test results. This discrepancy and banks’ inability to reconcile the differences makes it difficult to explain the results to the market. AFME supports improving the transparency of the constraints and overlays applied to the bank specific results and of the top-down models developed by the EBA and the ECB to calculate (e.g. fees and commissions) or challenge banks’ projections.

• AFME welcomes the expected revisions to the EBA methodology, and is willing to collaborate with the EBA to develop the roadmap to 2025 stress tests. In this context, AFME is particularly concerned about the severe treatment of capital markets activities in the EBA stress test methodology and scenarios vis-à-vis more traditional commercial and retail banking products.  This  creates a level playing field issue for banks that serve clients and finance the economy through these activities. AFME looks forward to discussing ways to address this level playing field issue with the EBA.

• We specifically  highlight concerns with the market risk methodology, suggesting a need for revising the prescribed floors on trading losses and addressing the lack of coherence and over-conservatism. The former in many cases replaces banks’ hard work on full revaluation of market risk under the scenario with a simple balance sheet multiplier, which is not risk sensitive. Beyond this particular case, the market risk methodology includes multiple shocks (full revaluation impact, increase in market liquidity reserves and loss of client revenues), which may be either mutually exclusive or overly conservative. In addition, the market risk methodology assumes that losses suffered in the first year of projections are never recovered in subsequent years, which is not aligned with the history of past market risk episodes.

• Finally, AFME and its members are keen on re-establishing of the process, with hopefully the number of ad-hoc templates reduced for future exercises. While we understand that there was an acute desire to capture data on topical items in direct response to the recent interest rate risk issues that resulted in bank failures in the US, it is difficult to manage the workload and commitments when additional templates are added late in the process. AFME recommends that the authorities should use where possible the alternate years when the EU-wide stress tests are not run to look at any additional targeted exercises outside the main perimeter of the EBA stress test.

ENDS -